Search Results for author: Jaime Spencer

Found 10 papers, 7 papers with code

Kick Back & Relax: Learning to Reconstruct the World by Watching SlowTV

1 code implementation ICCV 2023 Jaime Spencer, Chris Russell, Simon Hadfield, Richard Bowden

Unfortunately, existing approaches limit themselves to the automotive domain, resulting in models incapable of generalizing to complex environments such as natural or indoor settings.

Monocular Depth Estimation Motion Estimation +1

Deconstructing Self-Supervised Monocular Reconstruction: The Design Decisions that Matter

2 code implementations2 Aug 2022 Jaime Spencer, Chris Russell, Simon Hadfield, Richard Bowden

It is likely that many papers were not only optimized for particular datasets, but also for errors in the data and evaluation criteria.

Monocular Depth Estimation Monocular Reconstruction

Medusa: Universal Feature Learning via Attentional Multitasking

no code implementations12 Apr 2022 Jaime Spencer, Richard Bowden, Simon Hadfield

We argue that MTL is a stepping stone towards universal feature learning (UFL), which is the ability to learn generic features that can be applied to new tasks without retraining.

Multi-Task Learning

DeFeat-Net: General Monocular Depth via Simultaneous Unsupervised Representation Learning

1 code implementation CVPR 2020 Jaime Spencer, Richard Bowden, Simon Hadfield

In the current monocular depth research, the dominant approach is to employ unsupervised training on large datasets, driven by warped photometric consistency.

Monocular Depth Estimation Representation Learning

Same Features, Different Day: Weakly Supervised Feature Learning for Seasonal Invariance

1 code implementation CVPR 2020 Jaime Spencer, Richard Bowden, Simon Hadfield

The aim of this paper is to provide a dense feature representation that can be used to perform localization, sparse matching or image retrieval, regardless of the current seasonal or temporal appearance.

Image Retrieval Retrieval

Scale-Adaptive Neural Dense Features: Learning via Hierarchical Context Aggregation

1 code implementation CVPR 2019 Jaime Spencer, Richard Bowden, Simon Hadfield

In all cases, we show how incorporating SAND features results in better or comparable results to the baseline, whilst requiring little to no additional training.

Disparity Estimation Semantic Segmentation

Localisation via Deep Imagination: learn the features not the map

no code implementations19 Nov 2018 Jaime Spencer, Oscar Mendez, Richard Bowden, Simon Hadfield

In order to build the embedded map, we train a deep Siamese Fully Convolutional U-Net to perform dense feature extraction.

Cannot find the paper you are looking for? You can Submit a new open access paper.